1. Field of the Invention
The present invention concerns a particular construction that enables the fabrication of travelling wave tubes with very great bandwidth and very little dispersion. This construction employs supporting the helix of the delay line of a travelling wave tube by means of insulating dielectric supports placed between the helix and vanes or other metal supports projecting towards the center from a metal casing that surrounds the helix.
The invention also concerns a travelling wave tube fabricated according to this construction method.
2. Description of the Prior Art
Travelling wave tubes (TWTs) are well known in the prior art and are preferred to other microwave tubes for applications that require a very wide intrinsic passband in amplification. The wide passband permitted by the helical construction of the delay line results from the low dispersion of the electromagnetic waves that are propagated along the delay line as a function of the frequency: in other words, the velocity v of the wave that is propagated along the helical line depends only very little on the frequency of the wave in a wide range of frequencies centered on the nominal frequency of operation of the travelling wave tube.
The coupling between the high frequency (HF) signal applied to the input of the tube and, from there, to the helix delay line, on the one hand, and the electron beam, on the other, depends on the synchronism of the propagation of both of them along the longitudinal direction of the travelling wave tubes. The velocity of the electron beam depends, in an initial rough calculation, on the acceleration voltages created inside the tube and then, in a more precise calculation, it is modified by the exchange of energy that is produced with the electromagnetic field. While, in the initial rough calculation, the velocity of the high-frequency wave that is propagated along the helix depends only on the geometry of the helix, further calculation shows it also depends slightly on the frequency, and this finally restricts the passband of the travelling wave tube.
For practical reasons, the use of the travelling wave tube in amplifier equipment does not provide for the setting of the operating voltages to modify the velocity of the electrons of the beam when the signal to be amplified varies, and it is consequently desirable to have as small a variation as possible in the velocity of the electromagnetic wave as a function of frequency. However in all the methods of practical construction of a delay line, the phase velocity of the the electromagnetic wave depends on the frequency, and this gives rise to a dispersion of d=Vp/Vg, where Vp is the phase velocity along the delay line at the working frequency and Vg is the group velocity.
FIG. 1 reproduces a typical curve representing the variation of the ratio c/v as a function of the wavelength, where c is the velocity of light. It is applicable both to the prior art and to the present invention. A value of d approximately equal to unity means that the phase speed along the delay line is practically constant when the frequency varies: this is the condition to be fulfilled for wideband operation, if possible throughout the operating bandwidth. A very high value of d corresponds either to an infinite value of Vp (wave guided at the cut-off frequency) or to a value of Vg close to zero. This means that the energy is not propagated along the delay line.
In practice, for any wave propagation circuit (in this case the delay line) having periodic electrical or geometric characteristics, the circuit stops transmitting energy in a given mode of propagation at the frequencies such that the half wavelength in this mode is equal to the period of the geometrical characteristics of the delay line. These frequencies are called ".pi. mode cut-off frequencies". There are also zero mode cut-off frequencies when the phase difference on a period of the slow wave structure is equal to zero or to a multiple of 2.pi..
At these frequencies, the dispersion as defined above tends towards infinity. These cut-off frequencies cannot be avoided in a real physical circuit because Vg cannot grow indefinitely as and when Vp grows indefinitely, any more than V.sub.g can be prevented from becoming infinitesimally small for a finite value of Vp.
In the prior art, it is known that these cut-off frequencies can be shifted, but at the cost of a decrease in the efficiency of the travelling wave tube in operation: the intensity of the electrical field is reduced for a given level of the power being propagated in the delay line at a given phase velocity. This reduces the interaction with the electron beam in the tube. Should the cut-off frequencies be not shifted by any means, the experimentally observed cut-off frequencies may be called "natural cut-off frequencies".
According to one known prior art method for mitigating this drawback, an anisotropic load is added to the basic helix delay line. This gives a very low dispersion which may even become zero or negative.
The best known variant of this method for applying a load shown in FIG. 3, is the arrangement of U-shaped metal vanes with capacitive effect between the dielectric bars supporting the helix, the ends of which are positioned very near the helix (some tenths of a millimeter from it). It is difficult to obtain repeatable results economically in an industrial-scale fabrication process when this method is used. This method generally calls for resorting to a difficult brazing technique.
According to another known method in the prior art, capacitive loads are placed between the dielectric bars supporting the helix, as shown in FIG. 2, but this approach reduces the coupling impedance of the circuit and the efficiency of the tube.
In yet another known method, the above-mentioned U-shaped vanes, shown in FIG. 3, are replaced by localized metallizations of the dielectric bars supporting the helix, as shown in FIG. 4. This method is also difficult to implement on an industrial scale if repeatable results are to be obtained economically.
The invention is therefore aimed at obtaining a higher cut-off frequency without the drawbacks of prior art methods. The fundamental principles of physics used in the prior art can be brought out in a novel construction according to the invention. This gives a very low dispersion and, consequently, a widened useful passband while, at the same time, reducing the cost price of industrial-scale fabrication and the complexity of the helix assembly, and improving the repeatability of the characteristics of the tube.